The researchers aimed to improve how a computer model called the Storm Water Management Model (SWMM) predicts the quality of water flowing off green roofs. They tested SWMM's ability to estimate pollutants like total phosphorus (TP) and total nitrogen (TN) in runoff water from a green roof in Portland. Initially, SWMM's predictions were not accurate. By adjusting the model and simulating build-up and wash-off of pollutants, they improved the accurate prediction of TP but not TN. This work could help design better green roofs that release fewer pollutants like TP into the environment.